Multiperiod Traffic Congestion Estimation Model Based on an Urban Spatial Structure

With the rapid development of new urbanization, traffic congestion in Chinese cities has worsened, a problem not fully addressed by current models due to its complex link to urban spatial structure. This study proposes a Multiple-period Traffic Congestion Estimation Model using a random forest regression approach to predict congestion across three daily periods. The model achieves an average accuracy of over 85%, demonstrating significant predictive capability. By analyzing urban spatial factors such as road network structure and functional distribution, this study enhances our understanding of the interactions between urban spatial structures and traffic patterns. It offers practical value for optimizing urban traffic management and planning, while also promising environmental benefits through reduced congestion and emissions.